Questions and Answers :
Windows :
SETI Intel GPU computing process uses constant 10%+ CPU on a 8 core system
Message board moderation
Author | Message |
---|---|
tangyuhao Send message Joined: 3 Mar 12 Posts: 3 Credit: 1,968,862 RAC: 0 |
SETI Intel GPU computing process uses constant 10%+ CPU on a 8 core system. It leads my CPU always runs at highest frequency SETI@home v8 8.20 (opencl_intel_sah) Win10 1709 64bit Detailed log: 2017/10/20 1:23:25 | | cc_config.xml not found - using defaults 2017/10/20 1:23:25 | | Starting BOINC client version 7.8.3 for windows_x86_64 2017/10/20 1:23:25 | | log flags: file_xfer, sched_ops, task 2017/10/20 1:23:25 | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8 2017/10/20 1:23:25 | | Data directory: C:\ProgramData\BOINC 2017/10/20 1:23:25 | | Running under account TYH 2017/10/20 1:23:26 | | OpenCL: Intel GPU 0: Intel(R) HD Graphics 530 (driver version 22.20.16.4815, device version OpenCL 2.1, 13053MB, 13053MB available, 211 GFLOPS peak) 2017/10/20 1:23:26 | | OpenCL CPU: Intel(R) Core(TM) i7-6700T CPU @ 2.80GHz (OpenCL driver vendor: Intel(R) Corporation, driver version 7.5.0.550, device version OpenCL 2.0 (Build 550)) 2017/10/20 1:23:26 | | Host name: TYH-M900 2017/10/20 1:23:26 | | Processor: 8 GenuineIntel Intel(R) Core(TM) i7-6700T CPU @ 2.80GHz [Family 6 Model 94 Stepping 3] 2017/10/20 1:23:26 | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 vmx smx tm2 pbe fsgsbase bmi1 hle smep bmi2 2017/10/20 1:23:26 | | OS: Microsoft Windows 10: Education x64 Edition, (10.00.16299.00) 2017/10/20 1:23:26 | | Memory: 31.89 GB physical, 36.89 GB virtual 2017/10/20 1:23:26 | | Disk: 475.61 GB total, 320.71 GB free 2017/10/20 1:23:26 | | Local time is UTC +8 hours 2017/10/20 1:23:26 | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8365193; resource share 100 2017/10/20 1:23:26 | WUProp@Home | URL https://wuprop.boinc-af.org/; Computer ID 130866; resource share 100 2017/10/20 1:23:26 | | General prefs: from http://www.worldcommunitygrid.org/ (last modified 12-Feb-2017 06:30:14) 2017/10/20 1:23:26 | | Host location: none 2017/10/20 1:23:26 | | General prefs: using your defaults 2017/10/20 1:23:26 | | Reading preferences override file 2017/10/20 1:23:26 | | Preferences: 2017/10/20 1:23:26 | | max memory usage when active: 24490.30 MB 2017/10/20 1:23:26 | | max memory usage when idle: 29388.36 MB 2017/10/20 1:23:26 | | max disk usage: 5.00 GB 2017/10/20 1:23:26 | | don't compute while active 2017/10/20 1:23:26 | | don't use GPU while active 2017/10/20 1:23:26 | | suspend work if non-BOINC CPU load exceeds 25% 2017/10/20 1:23:26 | | (to change preferences, visit a project web site or select Preferences in the Manager) 2017/10/20 1:23:26 | | Using account manager BOINCstatsBAM! 2017/10/20 1:23:26 | | Suspending GPU computation - computer is in use 2017/10/20 1:29:35 | | Resuming GPU computation 2017/10/20 1:46:10 | SETI@home | Computation for task 29mr07ah.26730.15205.9.36.23_0 finished 2017/10/20 1:46:10 | SETI@home | Starting task blc04_2bit_guppi_57976_00991_HIP74666_0008.2741.0.22.45.77.vlar_1 2017/10/20 1:46:12 | SETI@home | Started upload of 29mr07ah.26730.15205.9.36.23_0_r756769269_0 2017/10/20 1:46:17 | SETI@home | Finished upload of 29mr07ah.26730.15205.9.36.23_0_r756769269_0 2017/10/20 1:46:17 | SETI@home | Sending scheduler request: To report completed tasks. 2017/10/20 1:46:17 | SETI@home | Reporting 1 completed tasks 2017/10/20 1:46:17 | SETI@home | Not requesting tasks: "no new tasks" requested via Manager 2017/10/20 1:46:20 | SETI@home | Scheduler request completed 2017/10/20 2:19:36 | SETI@home | Computation for task blc04_2bit_guppi_57976_00991_HIP74666_0008.2741.0.22.45.77.vlar_1 finished 2017/10/20 2:19:36 | SETI@home | Starting task 29mr07ah.11614.18477.14.41.66_1 2017/10/20 2:19:38 | SETI@home | Started upload of blc04_2bit_guppi_57976_00991_HIP74666_0008.2741.0.22.45.77.vlar_1_r1710038232_0 2017/10/20 2:19:43 | SETI@home | Finished upload of blc04_2bit_guppi_57976_00991_HIP74666_0008.2741.0.22.45.77.vlar_1_r1710038232_0 2017/10/20 2:19:47 | SETI@home | Sending scheduler request: To report completed tasks. 2017/10/20 2:19:47 | SETI@home | Reporting 1 completed tasks 2017/10/20 2:19:47 | SETI@home | Not requesting tasks: "no new tasks" requested via Manager 2017/10/20 2:19:50 | SETI@home | Scheduler request completed 2017/10/20 3:03:38 | SETI@home | Computation for task 29mr07ah.11614.18477.14.41.66_1 finished 2017/10/20 3:03:38 | SETI@home | Starting task 29mr07ah.11579.18477.13.40.173_0 2017/10/20 3:03:40 | SETI@home | Started upload of 29mr07ah.11614.18477.14.41.66_1_r1724216094_0 2017/10/20 3:03:43 | SETI@home | Finished upload of 29mr07ah.11614.18477.14.41.66_1_r1724216094_0 2017/10/20 3:03:43 | SETI@home | Sending scheduler request: To report completed tasks. 2017/10/20 3:03:43 | SETI@home | Reporting 1 completed tasks 2017/10/20 3:03:43 | SETI@home | Not requesting tasks: "no new tasks" requested via Manager 2017/10/20 3:03:45 | SETI@home | Scheduler request completed 2017/10/20 3:04:56 | | Suspending GPU computation - computer is in use 2017/10/20 3:23:02 | | Resuming GPU computation 2017/10/20 5:33:37 | SETI@home | Computation for task 29mr07ah.11579.18477.13.40.173_0 finished 2017/10/20 5:33:37 | SETI@home | Starting task 01ap07aa.18212.7434.10.37.210_1 2017/10/20 5:33:39 | SETI@home | Started upload of 29mr07ah.11579.18477.13.40.173_0_r681470442_0 2017/10/20 5:33:44 | SETI@home | Finished upload of 29mr07ah.11579.18477.13.40.173_0_r681470442_0 2017/10/20 5:33:47 | SETI@home | Sending scheduler request: To report completed tasks. 2017/10/20 5:33:47 | SETI@home | Reporting 1 completed tasks 2017/10/20 5:33:47 | SETI@home | Not requesting tasks: "no new tasks" requested via Manager 2017/10/20 5:33:49 | SETI@home | Scheduler request completed 2017/10/20 7:04:58 | SETI@home | Computation for task 01ap07aa.18212.7434.10.37.210_1 finished 2017/10/20 7:04:58 | SETI@home | Starting task 01ap07aa.18212.7434.10.37.124_1 2017/10/20 7:05:00 | SETI@home | Started upload of 01ap07aa.18212.7434.10.37.210_1_r2094799854_0 2017/10/20 7:05:03 | SETI@home | Finished upload of 01ap07aa.18212.7434.10.37.210_1_r2094799854_0 2017/10/20 7:05:06 | SETI@home | Sending scheduler request: To report completed tasks. 2017/10/20 7:05:06 | SETI@home | Reporting 1 completed tasks 2017/10/20 7:05:06 | SETI@home | Not requesting tasks: "no new tasks" requested via Manager 2017/10/20 7:05:09 | SETI@home | Scheduler request completed 2017/10/20 7:36:27 | | Suspending GPU computation - computer is in use 2017/10/20 7:41:28 | | Resuming GPU computation 2017/10/20 7:46:15 | | Suspending GPU computation - computer is in use |
tangyuhao Send message Joined: 3 Mar 12 Posts: 3 Credit: 1,968,862 RAC: 0 |
Anyone know why it happens and how to solve it, since it should cost less than 1% CPU while computing |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
2017/10/20 1:23:26 | | General prefs: using your defaults This isn't happening because of CPU use. It's happening because you have your preferences set to not do any crunching on the CPU or GPU while the computer is in use. In use means if a key is pressed on the keyboard or the mouse is moved. I've bolded the pertinent lines in your event log. If you want to crunch all the time, you'll need to uncheck "Suspend when computer is in use" and "Suspend GPU computing when computer is in use". |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
SETI Intel GPU computing process uses constant 10%+ CPU on a 8 core system. It leads my CPU always runs at highest frequencyThe Intel GPU is built into your CPU, and will therefore when used for calculations always use a significant part of the CPU. This can lead to extra heat, leading to extra noise from the cooling fans. In all cases the GPU is given data by the CPU, as the CPU does all the reading of data from disk, translating it into something the GPU can deal with, transporting it to the GPU, transporting the resulting data back to memory, translating it to previous data formats and writing it to disk. It is therefore a good idea to always keep one CPU core free so it is able to do all that, and whatever chores the OS and user has for it (OS house cleaning, opening programs, etc.) You can do this by changing the value for Use at most N% of the CPUs in your computing preferences to 87.5% (100 / 8 = 12.5 x 7). (Seeing that you're using the BAM account manager, you may have to change it there). |
tangyuhao Send message Joined: 3 Mar 12 Posts: 3 Credit: 1,968,862 RAC: 0 |
I understand my settings and do not run any task when i'm using my computer is what i want. Also, I know Intel GPU is a part of CPU, but they are still different parts. if only my GPU is full load, the CPU can still keep lower than 2GHz, but my situation is the GPU task uses much more CPU than it should be, and all my cores keeps 3GHz+ (I can manually limit my CPU max frequency, but it need my change it manually between I use computer and BOINC use computer) The following is some tasks on two of my computers, 8328949 is the computer behave normally, for example on the first line, 3,965.72 is running time, and 212.17 is the CPU time, it did not consume much CPU since it is a GPU task. But on the other machine 8365193, it goes wrong, for example third line, a GPU task with 7,901.28running time has 2,642.55 CPU time, this is unusual. 6103859039 2716201603 8328949 19 Oct 2017, 12:16:55 UTC 19 Oct 2017, 17:08:53 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,965.72 212.47 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103859040 2716201609 8328949 19 Oct 2017, 12:16:55 UTC 19 Oct 2017, 16:02:47 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,204.79 123.47 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103582076 2716069009 8365193 19 Oct 2017, 9:30:32 UTC 19 Oct 2017, 21:33:50 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 7,901.28 2,642.55 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103392438 2715978090 8328949 19 Oct 2017, 7:30:36 UTC 19 Oct 2017, 12:51:27 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 4,348.64 305.77 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103196869 2715884345 8328949 19 Oct 2017, 5:28:44 UTC 19 Oct 2017, 10:26:25 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,468.79 140.72 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103140140 2715857478 8365193 19 Oct 2017, 4:52:21 UTC 19 Oct 2017, 8:03:12 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 2,432.71 1,615.50 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103139887 2715857506 8365193 19 Oct 2017, 4:52:21 UTC 19 Oct 2017, 6:52:08 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 2,972.92 1,964.58 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6103139410 2715857080 8365193 19 Oct 2017, 4:52:21 UTC 19 Oct 2017, 12:22:43 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 2,829.92 1,901.95 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6101653231 2715138508 8328949 18 Oct 2017, 13:30:12 UTC 18 Oct 2017, 22:38:42 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,809.75 178.00 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6100984091 2714814242 8328949 18 Oct 2017, 7:10:09 UTC 18 Oct 2017, 13:30:12 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,850.42 232.02 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6100984093 2714814248 8328949 18 Oct 2017, 7:10:09 UTC 18 Oct 2017, 15:02:27 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 5,494.51 195.08 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6096993025 2712919551 8328949 16 Oct 2017, 10:43:54 UTC 18 Oct 2017, 3:48:27 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 4,029.22 291.30 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6092514052 2710783898 8328949 14 Oct 2017, 11:36:56 UTC 14 Oct 2017, 21:23:07 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 5,190.55 148.22 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6087902225 2708587698 8328949 12 Oct 2017, 2:21:06 UTC 12 Oct 2017, 10:45:11 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 12,644.57 442.97 ç‰å¾…ä¸ AstroPulse v7 v7.09 (opencl_intel_gpu_102) windows_intelx86 6078408253 2704043218 8328949 8 Oct 2017, 3:13:46 UTC 8 Oct 2017, 13:19:29 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 5,077.28 718.77 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6077987176 2703841873 8328949 7 Oct 2017, 22:29:29 UTC 8 Oct 2017, 9:59:06 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 19.18 12.38 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6077063134 2703400018 8328949 7 Oct 2017, 12:27:04 UTC 7 Oct 2017, 22:29:29 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 2,706.12 327.64 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6075348209 2702580632 8328949 6 Oct 2017, 18:33:40 UTC 7 Oct 2017, 2:51:30 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 4,564.74 239.89 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6072102620 2701034411 8328949 5 Oct 2017, 10:26:51 UTC 5 Oct 2017, 21:41:04 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 6,599.46 588.27 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 6045988011 2688527357 8328949 24 Sep 2017, 22:17:33 UTC 25 Sep 2017, 12:16:35 UTC 已完æˆï¼Œç‰å¾…éªŒè¯ 3,673.06 219.69 ç‰å¾…ä¸ SETI@home v8 v8.20 (opencl_intel_gpu_sah) windows_intelx86 |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
But on the other machine 8365193, it goes wrong, for example third line, a GPU task with 7,901.28running time has 2,642.55 CPU time, this is unusual.Seti's tasks are not all the same length. It depends on their angle range (the angle across the horizon at which they were recorded) how long they take. And even in that there can be differences. But let's take your other computer as an example as well: Task 6101446530 ran 2,228.68 seconds on the GPU, 100.17 seconds on the CPU. Task 6100984093 ran 5,494.51 seconds on the GPU, 195.08 seconds on the CPU. Task 6072102620 ran 6,599.46 seconds on the GPU, 588.27 seconds on the CPU. So the second task ran more than double and the third about triple the amount of time that the first did, just showing that there's different length tasks around that run at different length times. |
©2025 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.